Numerical modelling of scour erosion in the waterway constriction place by using SSIIM model

The process of scouring is removal of sediments from river bed, caused due to moving water or waves. Scour can be classified into two broad categories: General scour and local scour. Contraction scour, as a type of general scour, occurs due to a reduction in channel cross-sectional area. Local scour on the other hand occurs due to the direct effect of an obstruction on the flow field. Contraction scour is observed where the flow is constricted due to the placement of structures like bridges etc. The flow accelerates in constrictions which increase the bed shear stress and the turbulence associated with it. The development of contraction scour is noted when the critical shear stress of the bed materials is overcome by the bed shear stress. This paper provides a comprehensive discussion on scour within channel contractions. Scour within a channel contraction occurs where the waterway is restricted by the foundations or road approach embankments of a bridge. Alternately, contraction scour takes place when a bridge is located at a natural contraction in the river width. Contraction scour is an important phenomenon in rivers. To reach the aims of this work, both experimental and numerical modeling has been done. In this study, the experiments curried out in a flume with length of 20 m and width of 0.6 m, located in hydraulic laboratory of Shahrekord University. The average size of the sediments in the flume was d50 = 0.78 mm. The bed sediment layer was 16 cm thick. The experiments were run with three different discharges: 20, 30 and 40 l/s and flow depth was regulated at 30 cm. Also, the experiments were conducted with four different contractions of 20 cm, 30 cm 40 and 50 cm wide, which was resulted in three different the contraction ratio of 0.33, 0.5 and 0.67. SSIIM computer model has been used for contraction scour prediction in stream for two different conditions, including sudden (with angle of 90o) and gradual (with angle of 50o) contractions. Calibration of SSIIM model was conducted based on the results from the experimental study. SSIIM numerical model is capable of solving the Navier-Stokes equations on three-dimensional network with k-ε turbulence model. Likewise, it utilizes volume control with power upstream or second-order law for discretizing equations. This software also addresses to solve the passive pressure field using iteration technique and simple method (semi-implicit method for solving equations with pressure). After calibration of experiments to analyze the output results, Tecplot software was employed because the display of the obtained results from scouring through SSIIM was not clear. In current study, parameters such as flow pattern, scouring pattern, and water depth were simulated. Calibration of results was carried out by the experimental data. With monitoring model results, the start of scour for all models in the beginning of the narrowing part and the side walls was simulated. The greatest value of the scouring achieved at the same place as in experimental model. According to the results of the model regarding flow pattern simulation, it can be stated that flowing lines are constricted by reaching to narrowing part beginning and along the narrowing section changed to parallel lines. At the end of constriction, flow lines are being expanded and at the gradual opening, rotational currents are formed. Maximum amount of scour was calculated 12 cm in flow rate of 40 L/s and 20 cm constriction with 9.09% error that this amount was measured 11 cm in the laboratory. In addition, the least amount of scour was measured 1.8 cm in flow rate of 20 L/s and 40 cm constriction with 20% error in numerical model and 1.5 cm in the laboratory. Obtained results showed that the model was adequate for modeling of contraction scour, in parameters of scouring maximum depth, erosion pattern and circular flows. Accuracy of the model was acceptable according to the average of Nash – Sutcliffe coefficient of 0.923, average error of 10.18 and the maximum error of 20%. Based on the calibrated model, contraction scour was simulated for sudden transition and results showed that scour maximum depth was increased at a rate of 29.5%, comparing with gradual transition. This indicated that gradual transition plays a positive role to reducing contraction scour.